Issues

• [Bug] resource manager: capacity gained by a larger DAX device is lost when stale pool metadata is loaded

  bug
  xcena-dev/maru#103 · Filter by author seohui-XCENA opened on Oct 14, 2026
  1

  

• feat(maru_vllm): support models with multiple KV cache groups (e.g. Qwen3.5+)

  enhancement
  xcena-dev/maru#95 · Filter by author hhoikoo opened on Oct 1, 2026
  

• [RFC] Extend the native vLLM connector with CPU/CXL L1 and SSD L2 caching

  enhancement
  xcena-dev/maru#87 · Filter by author jooho-XCENA opened on Sep 23, 2026

• feat(maru_vllm): place the layer-overlap load with single_layer_kv_transfer instead of PyTorch indexing

  enhancement
  xcena-dev/maru#86 · Filter by author youngrok-XCENA opened on Sep 22, 2026

• feat(maru_vllm): maru-native CUDA kernel for packed GPU↔CXL KV transfers

  xcena-dev/maru#74 · Filter by author seohui-XCENA opened on Aug 19, 2026

• Defense-in-depth: bounds-check server-supplied offset/length before device ops in handler plugins

  xcena-dev/maru#59 · Filter by author hyunyul-XCENA opened on Jul 6, 2026

• prepare pypi organization

  xcena-dev/maru#52 · Filter by author darjeeling opened on May 28, 2026
  1

• feat: event-driven CXL device hotplug monitoring

  xcena-dev/maru#49 · Filter by author seohui-XCENA opened on Apr 17, 2026

• [Feature] Implement eviction and capacity reclaim on CXL pool exhaustion

  high-priority
  Feature

  xcena-dev/maru#4 · Filter by author seohui-XCENA opened on Mar 6, 2026